Master Linux Performance in 60 Seconds: 10 Essential Commands

When you notice performance problems on a Linux server, the first 60 seconds are crucial. Netflix’s performance engineering team recommends using a set of standard Linux command‑line tools to quickly assess system health.

The following ten commands provide a snapshot of overall system operation and per‑process resource usage.

1. uptime

$ uptime

This command shows the system’s average load over the past 1, 5, and 15 minutes, indicating how many tasks are waiting to run. A sudden jump in the 1‑minute load compared to the 15‑minute load often signals a recent issue.

2. dmesg | tail

$ dmesg | tail

Displays the most recent kernel messages, helping you spot errors such as OOM kills or TCP problems that may affect performance.

3. vmstat 1

$ vmstat 1

Shows virtual memory and CPU statistics refreshed every second. Key fields include:

r : runnable tasks (higher than CPU count indicates saturation).

free : free memory in KB.

si/so : pages swapped in/out (non‑zero means memory pressure).

us, sy, id, wa, st : CPU time spent in user, system, idle, I/O wait, and steal (virtualization) states.

High

wa

suggests I/O bottlenecks; high

sy

(>20%) may indicate inefficient kernel handling.

4. mpstat -P ALL 1

$ mpstat -P ALL 1

Displays per‑CPU utilization, useful for spotting uneven load that could point to single‑threaded workloads.

5. pidstat 1

$ pidstat 1

Shows per‑process CPU usage at one‑second intervals. The

%CPU

column is relative to all CPUs, so values above 100% indicate multi‑CPU consumption.

6. iostat -xz 1

$ iostat -xz 1

Provides block‑device statistics. Important metrics:

r/s, w/s, rkB/s, wkB/s : read/write request rates and data throughput.

await : average I/O request latency (high values signal saturation).

avgqu‑sz : average queue length (values >1 indicate a bottleneck).

%util : device utilization percentage; >60% often correlates with performance problems.

7. free -m

$ free -m

Shows memory usage. Pay attention to the

buffers

and

cached

columns; the

-/+ buffers/cache

line gives a more accurate view of memory actually used by applications.

8. sar -n DEV 1

$ sar -n DEV 1

Monitors network interface throughput (

rxkB/s

,

txkB/s

) and utilization (

%ifutil

), helping you determine if network bandwidth is a limiting factor.

9. sar -n TCP,ETCP 1

$ sar -n TCP,ETCP 1

Summarizes TCP activity, including new connections (

active/s

), accepted connections (

passive/s

), and retransmissions (

retrans/s

), which can indicate network or server load issues.

10. top

$ top

Combines many of the above metrics in a dynamic view, showing per‑process CPU and memory usage as well as overall system load. While useful, it refreshes quickly, so for intermittent problems tools like

vmstat

or

pidstat

that provide rolling output may be more reliable.